1USER_NAMESPACES(7) Linux Programmer's Manual USER_NAMESPACES(7)
2
6 user_namespaces - overview of Linux user namespaces
7
9 For an overview of namespaces, see namespaces(7).
10 User namespaces isolate security-related identifiers and attributes, in
12 particular, user IDs and group IDs (see credentials(7)), the root
13 directory, keys (see keyrings(7)), and capabilities (see capabili‐
14 ties(7)). A process's user and group IDs can be different inside and
15 outside a user namespace. In particular, a process can have a normal
16 unprivileged user ID outside a user namespace while at the same time
17 having a user ID of 0 inside the namespace; in other words, the process
18 has full privileges for operations inside the user namespace, but is
19 unprivileged for operations outside the namespace.
20 Nested namespaces, namespace membership
22 User namespaces can be nested; that is, each user namespace—except the
23 initial ("root") namespace—has a parent user namespace, and can have
24 zero or more child user namespaces. The parent user namespace is the
25 user namespace of the process that creates the user namespace via a
26 call to unshare(2) or clone(2) with the CLONE_NEWUSER flag.
27 The kernel imposes (since version 3.11) a limit of 32 nested levels of
29 user namespaces. Calls to unshare(2) or clone(2) that would cause this
30 limit to be exceeded fail with the error EUSERS.
31 Each process is a member of exactly one user namespace. A process cre‐
33 ated via fork(2) or clone(2) without the CLONE_NEWUSER flag is a member
34 of the same user namespace as its parent. A single-threaded process
35 can join another user namespace with setns(2) if it has the
36 CAP_SYS_ADMIN in that namespace; upon doing so, it gains a full set of
37 capabilities in that namespace.
38 A call to clone(2) or unshare(2) with the CLONE_NEWUSER flag makes the
40 new child process (for clone(2)) or the caller (for unshare(2)) a mem‐
41 ber of the new user namespace created by the call.
42 The NS_GET_PARENT ioctl(2) operation can be used to discover the
44 parental relationship between user namespaces; see ioctl_ns(2).
45 Capabilities
47 The child process created by clone(2) with the CLONE_NEWUSER flag
48 starts out with a complete set of capabilities in the new user names‐
49 pace. Likewise, a process that creates a new user namespace using
50 unshare(2) or joins an existing user namespace using setns(2) gains a
51 full set of capabilities in that namespace. On the other hand, that
52 process has no capabilities in the parent (in the case of clone(2)) or
53 previous (in the case of unshare(2) and setns(2)) user namespace, even
54 if the new namespace is created or joined by the root user (i.e., a
55 process with user ID 0 in the root namespace).
56 Note that a call to execve(2) will cause a process's capabilities to be
58 recalculated in the usual way (see capabilities(7)). Consequently,
59 unless the process has a user ID of 0 within the namespace, or the exe‐
60 cutable file has a nonempty inheritable capabilities mask, the process
61 will lose all capabilities. See the discussion of user and group ID
62 mappings, below.
63 A call to clone(2), unshare(2), or setns(2) using the CLONE_NEWUSER
65 flag sets the "securebits" flags (see capabilities(7)) to their default
66 values (all flags disabled) in the child (for clone(2)) or caller (for
67 unshare(2), or setns(2)). Note that because the caller no longer has
68 capabilities in its original user namespace after a call to setns(2),
69 it is not possible for a process to reset its "securebits" flags while
70 retaining its user namespace membership by using a pair of setns(2)
71 calls to move to another user namespace and then return to its original
72 user namespace.
73 The rules for determining whether or not a process has a capability in
75 a particular user namespace are as follows:
76 1. A process has a capability inside a user namespace if it is a member
78 of that namespace and it has the capability in its effective capa‐
79 bility set. A process can gain capabilities in its effective capa‐
80 bility set in various ways. For example, it may execute a set-user-
81 ID program or an executable with associated file capabilities. In
82 addition, a process may gain capabilities via the effect of
83 clone(2), unshare(2), or setns(2), as already described.
84 2. If a process has a capability in a user namespace, then it has that
86 capability in all child (and further removed descendant) namespaces
87 as well.
88 3. When a user namespace is created, the kernel records the effective
90 user ID of the creating process as being the "owner" of the names‐
91 pace. A process that resides in the parent of the user namespace
92 and whose effective user ID matches the owner of the namespace has
93 all capabilities in the namespace. By virtue of the previous rule,
94 this means that the process has all capabilities in all further
95 removed descendant user namespaces as well. The NS_GET_OWNER_UID
96 ioctl(2) operation can be used to discover the user ID of the owner
97 of the namespace; see ioctl_ns(2).
98 Effect of capabilities within a user namespace
100 Having a capability inside a user namespace permits a process to per‐
101 form operations (that require privilege) only on resources governed by
102 that namespace. In other words, having a capability in a user names‐
103 pace permits a process to perform privileged operations on resources
104 that are governed by (nonuser) namespaces associated with the user
105 namespace (see the next subsection).
106 On the other hand, there are many privileged operations that affect
108 resources that are not associated with any namespace type, for example,
109 changing the system time (governed by CAP_SYS_TIME), loading a kernel
110 module (governed by CAP_SYS_MODULE), and creating a device (governed by
111 CAP_MKNOD). Only a process with privileges in the initial user names‐
112 pace can perform such operations.
113 Holding CAP_SYS_ADMIN within the user namespace associated with a
115 process's mount namespace allows that process to create bind mounts and
116 mount the following types of filesystems:
117 * /proc (since Linux 3.8)
119 * /sys (since Linux 3.8)
120 * devpts (since Linux 3.9)
121 * tmpfs(5) (since Linux 3.9)
122 * ramfs (since Linux 3.9)
123 * mqueue (since Linux 3.9)
124 * bpf (since Linux 4.4)
125 Holding CAP_SYS_ADMIN within the user namespace associated with a
127 process's cgroup namespace allows (since Linux 4.6) that process to the
128 mount the cgroup version 2 filesystem and cgroup version 1 named hier‐
129 archies (i.e., cgroup filesystems mounted with the "none,name="
130 option).
131 Holding CAP_SYS_ADMIN within the user namespace associated with a
133 process's PID namespace allows (since Linux 3.8) that process to mount
134 /proc filesystems.
135 Note however, that mounting block-based filesystems can be done only by
137 a process that holds CAP_SYS_ADMIN in the initial user namespace.
138 Interaction of user namespaces and other types of namespaces
140 Starting in Linux 3.8, unprivileged processes can create user names‐
141 paces, and other the other types of namespaces can be created with just
142 the CAP_SYS_ADMIN capability in the caller's user namespace.
143 When a non-user-namespace is created, it is owned by the user namespace
145 in which the creating process was a member at the time of the creation
146 of the namespace. Actions on the non-user-namespace require capabili‐
147 ties in the corresponding user namespace.
148 If CLONE_NEWUSER is specified along with other CLONE_NEW* flags in a
150 single clone(2) or unshare(2) call, the user namespace is guaranteed to
151 be created first, giving the child (clone(2)) or caller (unshare(2))
152 privileges over the remaining namespaces created by the call. Thus, it
153 is possible for an unprivileged caller to specify this combination of
154 flags.
155 When a new namespace (other than a user namespace) is created via
157 clone(2) or unshare(2), the kernel records the user namespace of the
158 creating process against the new namespace. (This association can't be
159 changed.) When a process in the new namespace subsequently performs
160 privileged operations that operate on global resources isolated by the
161 namespace, the permission checks are performed according to the
162 process's capabilities in the user namespace that the kernel associated
163 with the new namespace. For example, suppose that a process attempts
164 to change the hostname (sethostname(2)), a resource governed by the UTS
165 namespace. In this case, the kernel will determine which user names‐
166 pace is associated with the process's UTS namespace, and check whether
167 the process has the required capability (CAP_SYS_ADMIN) in that user
168 namespace.
169 The NS_GET_USERNS ioctl(2) operation can be used to discover the user
171 namespace with which a non-user namespace is associated; see
172 ioctl_ns(2).
173 User and group ID mappings: uid_map and gid_map
175 When a user namespace is created, it starts out without a mapping of
176 user IDs (group IDs) to the parent user namespace. The
177 /proc/[pid]/uid_map and /proc/[pid]/gid_map files (available since
178 Linux 3.5) expose the mappings for user and group IDs inside the user
179 namespace for the process pid. These files can be read to view the
180 mappings in a user namespace and written to (once) to define the map‐
181 pings.
182 The description in the following paragraphs explains the details for
184 uid_map; gid_map is exactly the same, but each instance of "user ID" is
185 replaced by "group ID".
186 The uid_map file exposes the mapping of user IDs from the user names‐
188 pace of the process pid to the user namespace of the process that
189 opened uid_map (but see a qualification to this point below). In other
190 words, processes that are in different user namespaces will potentially
191 see different values when reading from a particular uid_map file,
192 depending on the user ID mappings for the user namespaces of the read‐
193 ing processes.
194 Each line in the uid_map file specifies a 1-to-1 mapping of a range of
196 contiguous user IDs between two user namespaces. (When a user names‐
197 pace is first created, this file is empty.) The specification in each
198 line takes the form of three numbers delimited by white space. The
199 first two numbers specify the starting user ID in each of the two user
200 namespaces. The third number specifies the length of the mapped range.
201 In detail, the fields are interpreted as follows:
202 (1) The start of the range of user IDs in the user namespace of the
204 process pid.
205 (2) The start of the range of user IDs to which the user IDs specified
207 by field one map. How field two is interpreted depends on whether
208 the process that opened uid_map and the process pid are in the same
209 user namespace, as follows:
210 a) If the two processes are in different user namespaces: field two
212 is the start of a range of user IDs in the user namespace of the
213 process that opened uid_map.
214 b) If the two processes are in the same user namespace: field two
216 is the start of the range of user IDs in the parent user names‐
217 pace of the process pid. This case enables the opener of
218 uid_map (the common case here is opening /proc/self/uid_map) to
219 see the mapping of user IDs into the user namespace of the
220 process that created this user namespace.
221 (3) The length of the range of user IDs that is mapped between the two
223 user namespaces.
224 System calls that return user IDs (group IDs)—for example, getuid(2),
226 getgid(2), and the credential fields in the structure returned by
227 stat(2)—return the user ID (group ID) mapped into the caller's user
228 namespace.
229 When a process accesses a file, its user and group IDs are mapped into
231 the initial user namespace for the purpose of permission checking and
232 assigning IDs when creating a file. When a process retrieves file user
233 and group IDs via stat(2), the IDs are mapped in the opposite direc‐
234 tion, to produce values relative to the process user and group ID map‐
235 pings.
236 The initial user namespace has no parent namespace, but, for consis‐
238 tency, the kernel provides dummy user and group ID mapping files for
239 this namespace. Looking at the uid_map file (gid_map is the same) from
240 a shell in the initial namespace shows:
241 $ cat /proc/$$/uid_map
243 0 0 4294967295
244 This mapping tells us that the range starting at user ID 0 in this
246 namespace maps to a range starting at 0 in the (nonexistent) parent
247 namespace, and the length of the range is the largest 32-bit unsigned
248 integer. This leaves 4294967295 (the 32-bit signed -1 value) unmapped.
249 This is deliberate: (uid_t) -1 is used in several interfaces (e.g.,
250 setreuid(2)) as a way to specify "no user ID". Leaving (uid_t) -1
251 unmapped and unusable guarantees that there will be no confusion when
252 using these interfaces.
253 Defining user and group ID mappings: writing to uid_map and gid_map
255 After the creation of a new user namespace, the uid_map file of one of
256 the processes in the namespace may be written to once to define the
257 mapping of user IDs in the new user namespace. An attempt to write
258 more than once to a uid_map file in a user namespace fails with the
259 error EPERM. Similar rules apply for gid_map files.
260 The lines written to uid_map (gid_map) must conform to the following
262 rules:
263 * The three fields must be valid numbers, and the last field must be
265 greater than 0.
266 * Lines are terminated by newline characters.
268 * There is a limit on the number of lines in the file. In Linux 4.14
270 and earlier, this limit was (arbitrarily) set at 5 lines. Since
271 Linux 4.15, the limit is 340 lines. In addition, the number of
272 bytes written to the file must be less than the system page size,
273 and the write must be performed at the start of the file (i.e.,
274 lseek(2) and pwrite(2) can't be used to write to nonzero offsets in
275 the file).
276 * The range of user IDs (group IDs) specified in each line cannot
278 overlap with the ranges in any other lines. In the initial imple‐
279 mentation (Linux 3.8), this requirement was satisfied by a simplis‐
280 tic implementation that imposed the further requirement that the
281 values in both field 1 and field 2 of successive lines must be in
282 ascending numerical order, which prevented some otherwise valid maps
283 from being created. Linux 3.9 and later fix this limitation, allow‐
284 ing any valid set of nonoverlapping maps.
285 * At least one line must be written to the file.
287 Writes that violate the above rules fail with the error EINVAL.
289 In order for a process to write to the /proc/[pid]/uid_map
291 (/proc/[pid]/gid_map) file, all of the following requirements must be
292 met:
293 1. The writing process must have the CAP_SETUID (CAP_SETGID) capability
295 in the user namespace of the process pid.
296 2. The writing process must either be in the user namespace of the
298 process pid or be in the parent user namespace of the process pid.
299 3. The mapped user IDs (group IDs) must in turn have a mapping in the
301 parent user namespace.
302 4. One of the following two cases applies:
304 * Either the writing process has the CAP_SETUID (CAP_SETGID) capa‐
306 bility in the parent user namespace.
307 + No further restrictions apply: the process can make mappings
309 to arbitrary user IDs (group IDs) in the parent user names‐
310 pace.
311 * Or otherwise all of the following restrictions apply:
313 + The data written to uid_map (gid_map) must consist of a single
315 line that maps the writing process's effective user ID (group
316 ID) in the parent user namespace to a user ID (group ID) in
317 the user namespace.
318 + The writing process must have the same effective user ID as
320 the process that created the user namespace.
321 + In the case of gid_map, use of the setgroups(2) system call
323 must first be denied by writing "deny" to the /proc/[pid]/set‐
324 groups file (see below) before writing to gid_map.
325 Writes that violate the above rules fail with the error EPERM.
327 Interaction with system calls that change process UIDs or GIDs
329 In a user namespace where the uid_map file has not been written, the
330 system calls that change user IDs will fail. Similarly, if the gid_map
331 file has not been written, the system calls that change group IDs will
332 fail. After the uid_map and gid_map files have been written, only the
333 mapped values may be used in system calls that change user and group
334 IDs.
335 For user IDs, the relevant system calls include setuid(2), setfsuid(2),
337 setreuid(2), and setresuid(2). For group IDs, the relevant system
338 calls include setgid(2), setfsgid(2), setregid(2), setresgid(2), and
339 setgroups(2).
340 Writing "deny" to the /proc/[pid]/setgroups file before writing to
342 /proc/[pid]/gid_map will permanently disable setgroups(2) in a user
343 namespace and allow writing to /proc/[pid]/gid_map without having the
344 CAP_SETGID capability in the parent user namespace.
345 The /proc/[pid]/setgroups file
347 The /proc/[pid]/setgroups file displays the string "allow" if processes
348 in the user namespace that contains the process pid are permitted to
349 employ the setgroups(2) system call; it displays "deny" if setgroups(2)
350 is not permitted in that user namespace. Note that regardless of the
351 value in the /proc/[pid]/setgroups file (and regardless of the
352 process's capabilities), calls to setgroups(2) are also not permitted
353 if /proc/[pid]/gid_map has not yet been set.
354 A privileged process (one with the CAP_SYS_ADMIN capability in the
356 namespace) may write either of the strings "allow" or "deny" to this
357 file before writing a group ID mapping for this user namespace to the
358 file /proc/[pid]/gid_map. Writing the string "deny" prevents any
359 process in the user namespace from employing setgroups(2).
360 The essence of the restrictions described in the preceding paragraph is
362 that it is permitted to write to /proc/[pid]/setgroups only so long as
363 calling setgroups(2) is disallowed because /proc/[pid]gid_map has not
364 been set. This ensures that a process cannot transition from a state
365 where setgroups(2) is allowed to a state where setgroups(2) is denied;
366 a process can transition only from setgroups(2) being disallowed to
367 setgroups(2) being allowed.
368 The default value of this file in the initial user namespace is
370 "allow".
371 Once /proc/[pid]/gid_map has been written to (which has the effect of
373 enabling setgroups(2) in the user namespace), it is no longer possible
374 to disallow setgroups(2) by writing "deny" to /proc/[pid]/setgroups
375 (the write fails with the error EPERM).
376 A child user namespace inherits the /proc/[pid]/setgroups setting from
378 its parent.
379 If the setgroups file has the value "deny", then the setgroups(2) sys‐
381 tem call can't subsequently be reenabled (by writing "allow" to the
382 file) in this user namespace. (Attempts to do so fail with the error
383 EPERM.) This restriction also propagates down to all child user names‐
384 paces of this user namespace.
385 The /proc/[pid]/setgroups file was added in Linux 3.19, but was back‐
387 ported to many earlier stable kernel series, because it addresses a
388 security issue. The issue concerned files with permissions such as
389 "rwx---rwx". Such files give fewer permissions to "group" than they do
390 to "other". This means that dropping groups using setgroups(2) might
391 allow a process file access that it did not formerly have. Before the
392 existence of user namespaces this was not a concern, since only a priv‐
393 ileged process (one with the CAP_SETGID capability) could call set‐
394 groups(2). However, with the introduction of user namespaces, it
395 became possible for an unprivileged process to create a new namespace
396 in which the user had all privileges. This then allowed formerly
397 unprivileged users to drop groups and thus gain file access that they
398 did not previously have. The /proc/[pid]/setgroups file was added to
399 address this security issue, by denying any pathway for an unprivileged
400 process to drop groups with setgroups(2).
401 Unmapped user and group IDs
403 There are various places where an unmapped user ID (group ID) may be
404 exposed to user space. For example, the first process in a new user
405 namespace may call getuid(2) before a user ID mapping has been defined
406 for the namespace. In most such cases, an unmapped user ID is con‐
407 verted to the overflow user ID (group ID); the default value for the
408 overflow user ID (group ID) is 65534. See the descriptions of
409 /proc/sys/kernel/overflowuid and /proc/sys/kernel/overflowgid in
410 proc(5).
411 The cases where unmapped IDs are mapped in this fashion include system
413 calls that return user IDs (getuid(2), getgid(2), and similar), creden‐
414 tials passed over a UNIX domain socket, credentials returned by
415 stat(2), waitid(2), and the System V IPC "ctl" IPC_STAT operations,
416 credentials exposed by /proc/[pid]/status and the files in
417 /proc/sysvipc/*, credentials returned via the si_uid field in the sig‐
418 info_t received with a signal (see sigaction(2)), credentials written
419 to the process accounting file (see acct(5)), and credentials returned
420 with POSIX message queue notifications (see mq_notify(3)).
421 There is one notable case where unmapped user and group IDs are not
423 converted to the corresponding overflow ID value. When viewing a
424 uid_map or gid_map file in which there is no mapping for the second
425 field, that field is displayed as 4294967295 (-1 as an unsigned inte‐
426 ger).
427 Set-user-ID and set-group-ID programs
429 When a process inside a user namespace executes a set-user-ID (set-
430 group-ID) program, the process's effective user (group) ID inside the
431 namespace is changed to whatever value is mapped for the user (group)
432 ID of the file. However, if either the user or the group ID of the
433 file has no mapping inside the namespace, the set-user-ID (set-group-
434 ID) bit is silently ignored: the new program is executed, but the
435 process's effective user (group) ID is left unchanged. (This mirrors
436 the semantics of executing a set-user-ID or set-group-ID program that
437 resides on a filesystem that was mounted with the MS_NOSUID flag, as
438 described in mount(2).)
439 Miscellaneous
441 When a process's user and group IDs are passed over a UNIX domain
442 socket to a process in a different user namespace (see the description
443 of SCM_CREDENTIALS in unix(7)), they are translated into the corre‐
444 sponding values as per the receiving process's user and group ID map‐
445 pings.
446
448 Namespaces are a Linux-specific feature.
449
451 Over the years, there have been a lot of features that have been added
452 to the Linux kernel that have been made available only to privileged
453 users because of their potential to confuse set-user-ID-root applica‐
454 tions. In general, it becomes safe to allow the root user in a user
455 namespace to use those features because it is impossible, while in a
456 user namespace, to gain more privilege than the root user of a user
457 namespace has.
458 Availability
460 Use of user namespaces requires a kernel that is configured with the
461 CONFIG_USER_NS option. User namespaces require support in a range of
462 subsystems across the kernel. When an unsupported subsystem is config‐
463 ured into the kernel, it is not possible to configure user namespaces
464 support.
465 As at Linux 3.8, most relevant subsystems supported user namespaces,
467 but a number of filesystems did not have the infrastructure needed to
468 map user and group IDs between user namespaces. Linux 3.9 added the
469 required infrastructure support for many of the remaining unsupported
470 filesystems (Plan 9 (9P), Andrew File System (AFS), Ceph, CIFS, CODA,
471 NFS, and OCFS2). Linux 3.12 added support the last of the unsupported
472 major filesystems, XFS.
473
475 The program below is designed to allow experimenting with user names‐
476 paces, as well as other types of namespaces. It creates namespaces as
477 specified by command-line options and then executes a command inside
478 those namespaces. The comments and usage() function inside the program
479 provide a full explanation of the program. The following shell session
480 demonstrates its use.
481 First, we look at the run-time environment:
483 $ uname -rs # Need Linux 3.8 or later
485 Linux 3.8.0
486 $ id -u # Running as unprivileged user
487 1000
488 $ id -g
489 1000
490 Now start a new shell in new user (-U), mount (-m), and PID (-p) names‐
492 paces, with user ID (-M) and group ID (-G) 1000 mapped to 0 inside the
493 user namespace:
494 $ ./userns_child_exec -p -m -U -M '0 1000 1' -G '0 1000 1' bash
496 The shell has PID 1, because it is the first process in the new PID
498 namespace:
499 bash$ echo $$
501 1
502 Mounting a new /proc filesystem and listing all of the processes visi‐
503 ble in the new PID namespace shows that the shell can't see any pro‐
504 cesses outside the PID namespace:
505 bash$ mount -t proc proc /proc
507 bash$ ps ax
508 PID TTY STAT TIME COMMAND
509 1 pts/3 S 0:00 bash
510 22 pts/3 R+ 0:00 ps ax
511 Inside the user namespace, the shell has user and group ID 0, and a
513 full set of permitted and effective capabilities:
514 bash$ cat /proc/$$/status | egrep '^[UG]id'
516 Uid: 0 0 0 0
517 Gid: 0 0 0 0
518 bash$ cat /proc/$$/status | egrep '^Cap(Prm|Inh|Eff)'
519 CapInh: 0000000000000000
520 CapPrm: 0000001fffffffff
521 CapEff: 0000001fffffffff
522 Program source
524 /* userns_child_exec.c
526 Licensed under GNU General Public License v2 or later
528 Create a child process that executes a shell command in new
530 namespace(s); allow UID and GID mappings to be specified when
531 creating a user namespace.
532 */
533 #define _GNU_SOURCE
534 #include <sched.h>
535 #include <unistd.h>
536 #include <stdlib.h>
537 #include <sys/wait.h>
538 #include <signal.h>
539 #include <fcntl.h>
540 #include <stdio.h>
541 #include <string.h>
542 #include <limits.h>
543 #include <errno.h>
544 /* A simple error-handling function: print an error message based
546 on the value in 'errno' and terminate the calling process */
547 #define errExit(msg) do { perror(msg); exit(EXIT_FAILURE); \
549 } while (0)
550 struct child_args {
552 char **argv; /* Command to be executed by child, with args */
553 int pipe_fd[2]; /* Pipe used to synchronize parent and child */
554 };
555 static int verbose;
557 static void
559 usage(char *pname)
560 {
561 fprintf(stderr, "Usage: %s [options] cmd [arg...]\n\n", pname);
562 fprintf(stderr, "Create a child process that executes a shell "
563 "command in a new user namespace,\n"
564 "and possibly also other new namespace(s).\n\n");
565 fprintf(stderr, "Options can be:\n\n");
566 #define fpe(str) fprintf(stderr, " %s", str);
567 fpe("-i New IPC namespace\n");
568 fpe("-m New mount namespace\n");
569 fpe("-n New network namespace\n");
570 fpe("-p New PID namespace\n");
571 fpe("-u New UTS namespace\n");
572 fpe("-U New user namespace\n");
573 fpe("-M uid_map Specify UID map for user namespace\n");
574 fpe("-G gid_map Specify GID map for user namespace\n");
575 fpe("-z Map user's UID and GID to 0 in user namespace\n");
576 fpe(" (equivalent to: -M '0 <uid> 1' -G '0 <gid> 1')\n");
577 fpe("-v Display verbose messages\n");
578 fpe("\n");
579 fpe("If -z, -M, or -G is specified, -U is required.\n");
580 fpe("It is not permitted to specify both -z and either -M or -G.\n");
581 fpe("\n");
582 fpe("Map strings for -M and -G consist of records of the form:\n");
583 fpe("\n");
584 fpe(" ID-inside-ns ID-outside-ns len\n");
585 fpe("\n");
586 fpe("A map string can contain multiple records, separated"
587 " by commas;\n");
588 fpe("the commas are replaced by newlines before writing"
589 " to map files.\n");
590 exit(EXIT_FAILURE);
592 }
593 /* Update the mapping file 'map_file', with the value provided in
595 'mapping', a string that defines a UID or GID mapping. A UID or
596 GID mapping consists of one or more newline-delimited records
597 of the form:
598 ID_inside-ns ID-outside-ns length
600 Requiring the user to supply a string that contains newlines is
602 of course inconvenient for command-line use. Thus, we permit the
603 use of commas to delimit records in this string, and replace them
604 with newlines before writing the string to the file. */
605 static void
607 update_map(char *mapping, char *map_file)
608 {
609 int fd, j;
610 size_t map_len; /* Length of 'mapping' */
611 /* Replace commas in mapping string with newlines */
613 map_len = strlen(mapping);
615 for (j = 0; j < map_len; j++)
616 if (mapping[j] == ',')
617 mapping[j] = '\n';
618 fd = open(map_file, O_RDWR);
620 if (fd == -1) {
621 fprintf(stderr, "ERROR: open %s: %s\n", map_file,
622 strerror(errno));
623 exit(EXIT_FAILURE);
624 }
625 if (write(fd, mapping, map_len) != map_len) {
627 fprintf(stderr, "ERROR: write %s: %s\n", map_file,
628 strerror(errno));
629 exit(EXIT_FAILURE);
630 }
631 close(fd);
633 }
634 /* Linux 3.19 made a change in the handling of setgroups(2) and the
636 'gid_map' file to address a security issue. The issue allowed
637 *unprivileged* users to employ user namespaces in order to drop
638 The upshot of the 3.19 changes is that in order to update the
639 'gid_maps' file, use of the setgroups() system call in this
640 user namespace must first be disabled by writing "deny" to one of
641 the /proc/PID/setgroups files for this namespace. That is the
642 purpose of the following function. */
643 static void
645 proc_setgroups_write(pid_t child_pid, char *str)
646 {
647 char setgroups_path[PATH_MAX];
648 int fd;
649 snprintf(setgroups_path, PATH_MAX, "/proc/%ld/setgroups",
651 (long) child_pid);
652 fd = open(setgroups_path, O_RDWR);
654 if (fd == -1) {
655 /* We may be on a system that doesn't support
657 /proc/PID/setgroups. In that case, the file won't exist,
658 and the system won't impose the restrictions that Linux 3.19
659 added. That's fine: we don't need to do anything in order
660 to permit 'gid_map' to be updated.
661 However, if the error from open() was something other than
663 the ENOENT error that is expected for that case, let the
664 user know. */
665 if (errno != ENOENT)
667 fprintf(stderr, "ERROR: open %s: %s\n", setgroups_path,
668 strerror(errno));
669 return;
670 }
671 if (write(fd, str, strlen(str)) == -1)
673 fprintf(stderr, "ERROR: write %s: %s\n", setgroups_path,
674 strerror(errno));
675 close(fd);
677 }
678 static int /* Start function for cloned child */
680 childFunc(void *arg)
681 {
682 struct child_args *args = (struct child_args *) arg;
683 char ch;
684 /* Wait until the parent has updated the UID and GID mappings.
686 See the comment in main(). We wait for end of file on a
687 pipe that will be closed by the parent process once it has
688 updated the mappings. */
689 close(args->pipe_fd[1]); /* Close our descriptor for the write
691 end of the pipe so that we see EOF
692 when parent closes its descriptor */
693 if (read(args->pipe_fd[0], &ch, 1) != 0) {
694 fprintf(stderr,
695 "Failure in child: read from pipe returned != 0\n");
696 exit(EXIT_FAILURE);
697 }
698 close(args->pipe_fd[0]);
700 /* Execute a shell command */
702 printf("About to exec %s\n", args->argv[0]);
704 execvp(args->argv[0], args->argv);
705 errExit("execvp");
706 }
707 #define STACK_SIZE (1024 * 1024)
709 static char child_stack[STACK_SIZE]; /* Space for child's stack */
711 int
713 main(int argc, char *argv[])
714 {
715 int flags, opt, map_zero;
716 pid_t child_pid;
717 struct child_args args;
718 char *uid_map, *gid_map;
719 const int MAP_BUF_SIZE = 100;
720 char map_buf[MAP_BUF_SIZE];
721 char map_path[PATH_MAX];
722 /* Parse command-line options. The initial '+' character in
724 the final getopt() argument prevents GNU-style permutation
725 of command-line options. That's useful, since sometimes
726 the 'command' to be executed by this program itself
727 has command-line options. We don't want getopt() to treat
728 those as options to this program. */
729 flags = 0;
731 verbose = 0;
732 gid_map = NULL;
733 uid_map = NULL;
734 map_zero = 0;
735 while ((opt = getopt(argc, argv, "+imnpuUM:G:zv")) != -1) {
736 switch (opt) {
737 case 'i': flags |= CLONE_NEWIPC; break;
738 case 'm': flags |= CLONE_NEWNS; break;
739 case 'n': flags |= CLONE_NEWNET; break;
740 case 'p': flags |= CLONE_NEWPID; break;
741 case 'u': flags |= CLONE_NEWUTS; break;
742 case 'v': verbose = 1; break;
743 case 'z': map_zero = 1; break;
744 case 'M': uid_map = optarg; break;
745 case 'G': gid_map = optarg; break;
746 case 'U': flags |= CLONE_NEWUSER; break;
747 default: usage(argv[0]);
748 }
749 }
750 /* -M or -G without -U is nonsensical */
752 if (((uid_map != NULL || gid_map != NULL || map_zero) &&
754 !(flags & CLONE_NEWUSER)) ||
755 (map_zero && (uid_map != NULL || gid_map != NULL)))
756 usage(argv[0]);
757 args.argv = &argv[optind];
759 /* We use a pipe to synchronize the parent and child, in order to
761 ensure that the parent sets the UID and GID maps before the child
762 calls execve(). This ensures that the child maintains its
763 capabilities during the execve() in the common case where we
764 want to map the child's effective user ID to 0 in the new user
765 namespace. Without this synchronization, the child would lose
766 its capabilities if it performed an execve() with nonzero
767 user IDs (see the capabilities(7) man page for details of the
768 transformation of a process's capabilities during execve()). */
769 if (pipe(args.pipe_fd) == -1)
771 errExit("pipe");
772 /* Create the child in new namespace(s) */
774 child_pid = clone(childFunc, child_stack + STACK_SIZE,
776 flags | SIGCHLD, &args);
777 if (child_pid == -1)
778 errExit("clone");
779 /* Parent falls through to here */
781 if (verbose)
783 printf("%s: PID of child created by clone() is %ld\n",
784 argv[0], (long) child_pid);
785 /* Update the UID and GID maps in the child */
787 if (uid_map != NULL || map_zero) {
789 snprintf(map_path, PATH_MAX, "/proc/%ld/uid_map",
790 (long) child_pid);
791 if (map_zero) {
792 snprintf(map_buf, MAP_BUF_SIZE, "0 %ld 1", (long) getuid());
793 uid_map = map_buf;
794 }
795 update_map(uid_map, map_path);
796 }
797 if (gid_map != NULL || map_zero) {
799 proc_setgroups_write(child_pid, "deny");
800 snprintf(map_path, PATH_MAX, "/proc/%ld/gid_map",
802 (long) child_pid);
803 if (map_zero) {
804 snprintf(map_buf, MAP_BUF_SIZE, "0 %ld 1", (long) getgid());
805 gid_map = map_buf;
806 }
807 update_map(gid_map, map_path);
808 }
809 /* Close the write end of the pipe, to signal to the child that we
811 have updated the UID and GID maps */
812 close(args.pipe_fd[1]);
814 if (waitpid(child_pid, NULL, 0) == -1) /* Wait for child */
816 errExit("waitpid");
817 if (verbose)
819 printf("%s: terminating\n", argv[0]);
820 exit(EXIT_SUCCESS);
822 }
823
825 newgidmap(1), newuidmap(1), clone(2), ptrace(2), setns(2), unshare(2),
826 proc(5), subgid(5), subuid(5), capabilities(7), cgroup_namespaces(7)
827 credentials(7), namespaces(7), pid_namespaces(7)
828 The kernel source file Documentation/namespaces/resource-control.txt.
830
832 This page is part of release 4.16 of the Linux man-pages project. A
833 description of the project, information about reporting bugs, and the
834 latest version of this page, can be found at
835 https://www.kernel.org/doc/man-pages/.
836Linux 2018-02-02 USER_NAMESPACES(7)